Dr. Seok's Cancer-Cell Digital Twin, First Steps

Switch a cell off and on,
and turn cancer back

Treat one gene as one switch and a cell becomes a short string of 0s and 1s. Roll that string forward by the rules and out comes the cell's fate.
Over these 20 lessons you build an eight-gene cancer-cell digital twin with your own hands, then try every possible genetic intervention to find the one that erases the cancer attractor. It is the method by which Prof. Kwang-Hyun Cho's KAIST team found DDX54 (MObyDiCK), rebuilt by hand at the smallest scale that still works. Python runs right inside this page. There is nothing to install.

🐍 Python🧬 Boolean networks🌀 Attractors 💊 Virtual screening🔁 Feedback loops🎯 Cancer reversion🇬🇧 English

Start lesson 1 →

How you'll learn

🧱Built from the ground up

You start with a single switch. State vectors, logic rules, synchronous update, attractors and basins go on one layer at a time, until lesson 7 stands the finished twin up.

▶️You compute all of it

All 256 cells get rolled to the end, not one of them skipped. What is impossible at a lab bench takes a second in your browser.

⚠️How to fool yourself, first

How to smash the gauge and report 'zero cancer', how to cut the brake and create more of it, how to swap the denominator and inflate the effect — you step into each trap yourself.

🛠️Learning by handling it

One lab widget per lesson — press the switches, replay the trajectory, roll a marble down the valley, flip an arrow and watch what happens.

Full curriculum

A gene is a switch — how one cell becomes a handful of numbers

State vectors and 2ⁿ

Rules make the cell — three words: AND, OR, NOT

Writing regulation down as logic

Letting time run — what 'all at once' really means

Trajectories, and the trap that ordering sets

Attractors — how many fates does a cell have?

Fixed points, cycles, and 'a state already seen'

Basins — roll all 16 cells and draw the map of fates

What exhaustive enumeration buys you

Naming things — which attractor is 'cancer'?

A phenotype is an interpretation, not a computation

Standing the digital twin up — an eight-gene cancer cell

DDX54 · p53 · MDM2 · CD47 · PDL1 and the immune cell

Feedback loops — why do p53 and MDM2 oscillate?

The clock that negative feedback builds

Two valleys — why a cancer cell never comes back on its own

Bistability and the threshold you cannot recross

All at once vs one at a time — the update scheme changes the answer

Synchronous, asynchronous, and artefacts of the model

Perturbation — the one line that pins a gene down

Knockout (KO) and overexpression (OE)

The first virtual knockout — switching DDX54 off

The moment the whole landscape changes

Exhaustive screening — what MObyDiCK does

Switch every gene off, and on, one at a time

Reversion efficiency — what exactly are we counting?

The problem of choosing a denominator

Killing, reverting — and the third state

The 'dormancy' that combination therapy creates

Three traps — smashing the gauge, cutting the brake, the lure of zero

Three ways a model lies to you

Combinatorial explosion — double perturbations and the wall of 2ⁿ

Why exhaustive enumeration becomes impossible so quickly

What if the rules are wrong — shake one arrow and see

Robustness checks, and their limits

From prediction to experiment — what the twin can and cannot say

The gap between simulation and the wet lab

Wrapping up — what we built, and the real MObyDiCK

What twenty lessons leave you with